RCD tripping instantly despite nothing plugged in

[Iggifer]

What about a fault on another circuit that is manifesting itself when an unrelated circuit is energised?

THE worst kind of fault

 

[bernardgreen]

What about a Neutral to GROUND fault where the incoming Neutral is not at the same potential as the GROUND ( not the CPC "Earth" ) at the fault site.

 

[JohnW2]

What about a fault on another circuit that is manifesting itself when an unrelated circuit is energised?

I may be missing something but I'm struggling to think of how that could be happening in the OP's case. If the sockets circuit in question really is totally unloaded (including removal of the boiler's fuse), then I can't see how "energising the circuit" (switching it's MCB on) could affect any other circuit - it would merely be "energising" a pile of cables which were not connected to anything.

The only way I could see that switching on the (SP) MCB could cause an RCD to trip would be if there were some sort of L-E fault on that circuit. ... am I missing something?

I suppsoe that cross-connections between circuits might facilitate the sort of thing you are postulating, but that seems extremely improbable!

KInd Regards, John

 

[JohnW2]

What about a Neutral to GROUND fault where the incoming Neutral is not at the same potential as the GROUND ( not the CPC "Earth" ) at the fault site.

I am struggling to see why/how, in terms of RCD operation, that would be any different from a N-CPC fault. All an RCD knows is that there is an L-N imbalance and, if such an imbalance exists, it neither knows nor cares where the 'missing neutral current' has gone. Perhaps you could explain what you are thinking of?

Kind Regards, John

 

[bernardgreen]

 

[JohnW2]

<diagram>

Yes, but I was talking in the context of the OP's situation, not 'in general'. What you describe would be a 'permanent' situation, which would result in the RCD tripping whenever there was sufficient total load on the local network (even if there were no load within the actual installation). AIUI, in the OP's case, the RCD trips don't occur if he isolates the L of one of his sockets circuits (with no loads connected to that circuit) - so I don't see how the mechanism you describe could be responsible.

Kind Regards, John

 

[bernardgreen]

all seems fine apart from some Cu corrosion on the wires leading to a socket on a damp wall.

Which could be creating a Neutral to Ground fault. If you move point B to the consumer unit neutral bar then potential difference A to B created by the Neutral current from one appliance could be enough to have a 30 mA ground current.

My point is when looking for faults that are tripping an RCD keep an open mind about what the possible causes are.

 

[JohnW2]

My point is when looking for faults that are tripping an RCD keep an open mind about what the possible causes are.

Indeed - but if (as apparently in the OP's case), opening a single-pole MCB on one (totally unloaded) circuit prevents the RCD trips, that excludes a lot of the hypothetical things that could be causing the problem if that were not the case. In particular, AFAICS, it excludes any N-cpc or N-ground fault as the cause, doesn't it?

Kind Regards, John

 

[GaryGarden]

Nothing changed on the circuit at all, nor have I hammered in any nails around the sockets! Just tested after disconnecting the faceplate entirely with the same

But have you hammered any nails in or done any drilling elsewhere? Cables have been known to take many weird and wounderful routes.

 

[ericmark]

Removing a fuse does not disconnect neutral. In the main the switch on a fused connection unit (FCU) does so ensure it is switched off. Although rare RCD's do go faulty, the problem is without test gear one can go around in circles looking for a fault.

I had a so called electrician fit a new consumer unit for my dad. I arrived at the house to find him with no test gears trying to find out why the RCD would not hold in. He had disconnected the ring and without that connected it would work OK. He tried a replacement RCD which improved matters but still tripped.

Next day I brought the test gear to dads house and started to test. Found both RCD's passed but one clearly far more sensitive than the other. Fault was some old PVC cable which had over heated causing the plasticizer to leach out and then insulation had cracked. Even with full test set this took some finding. With out the test gear that electrician would have never found the fault.

Problem with test gear used incorrectly it is dangerous so even if you hired a test set it would not really be the way to go. Even hiring is expensive as needs the calibration testing after each hire so when I looked at hiring it was £75 per week with week minimum hire. To buy around £750 so I would say likely both saver and cheaper not to DIY this work.

Once basics have been done which it would seem you have done then time to get an electrician with his test set to find the fault.

 

[JohnW2]

Removing a fuse does not disconnect neutral. In the main the switch on a fused connection unit (FCU) does so ensure it is switched off.

That's true - but in the OP's case it appears that switch off the (presumably single-pole) MCB for the circuit stops the RCD trips from occurring, which would not normally be the case if there were a N-E fault.

Kind Regards, John

 

[ericmark]

I seem to remember it was you who pointed out how load will cause neutral - earth faults to trip a RCD although in this case he does say no load.

 

[JohnW2]

I seem to remember it was you who pointed out how load will cause neutral - earth faults to trip a RCD although in this case he does say no load.

That's the point - with everything unplugged and the boiler fuse removed (and maybe its FCU switched off), switching on the circuit's MCB should not result in any increase in load - so shouldn't result in a trip even if there were an N-E fault.

Kind Regards, John

 

[ericmark]

I seem to remember it was you who pointed out how load will cause neutral - earth faults to trip a RCD although in this case he does say no load.

That's the point - with everything unplugged and the boiler fuse removed (and maybe its FCU switched off), switching on the circuit's MCB should not result in any increase in load - so shouldn't result in a trip even if there were an N-E fault.

Kind Regards, John

Assuming TN-C-S supply. Didn't think at time of first post but with a TT supply then earth neutral fault could trip RCD with no load.

 

[bernardgreen]

Assuming TN-C-S supply. Didn't think at time of first post but with a TT supply then earth neutral fault could trip RCD with no load.

Which is the same as Neutral to GROUND ( Not Earth ) fault on a TN-C-S system. The potential that forces a current on the Neutral through the RCD sensor coil is the potential difference between the incoming Neutral and the local Ground.